The Credibility Factor: Why a Great Claims Year Barely Moves Your Renewal

Your claims ran low all year. Fewer big hospital stays, a calmer specialty drug bill, a workforce that mostly stayed healthy. Then the renewal landed and the number still climbed. That gap between the year you had and the price you were quoted has a name in the actuary's office, and almost nobody at the carrier says it out loud.

The renewal that ignored your good year

Start with the number every mid-market employer felt in 2025. The KFF Employer Health Benefits Survey put the average family premium at $26,993, up 6% from the year before. Single coverage hit $9,325, up 5%. Stretch the lens to five years and family premiums are up 26%.

Those are averages across employers who had bad years and employers who had good ones. So here is the puzzle. If your group ran below its expected claims, why did your renewal look so much like the group down the street that got hammered by a premature birth and two cancer cases? You did everything right. The market moved anyway.

The answer is not that your carrier ignored your data. The answer is that it only partly believed it.

Timing sharpens the sting. Carriers price off an experience period, usually about twelve months of claims ending a few months before your renewal date, with a run-out window to catch bills that arrive late. A strong stretch that showed up recently may sit only partly inside that window. So even the portion of your good year the carrier is willing to credit can get clipped by the calendar before credibility ever touches it.

The word your carrier never says out loud: credibility

Every fully-insured or experience-rated renewal is a blend of two numbers. The first is your own group's recent claims experience, the actual dollars your people spent. The second is a pooled "manual" rate, a baseline the carrier builds from thousands of other groups that look demographically like yours.

A credibility factor decides how much of each number goes into the final price. Call it Z. It runs from 0 to 1. When Z is high, the carrier trusts your experience and prices mostly off your year. When Z is low, it leans on the manual rate and your good year barely registers.

Think of it as a dial, not a switch. A group with Z of 0.30 sees its renewal built from 30% its own numbers and 70% the pooled baseline. Your quiet claims year is real. The carrier just weights it at thirty cents on the dollar and fills the rest with the crowd. That is why a great year can shave far less off your renewal than the year itself deserved.

Baseball makes the logic obvious. A hitter who goes 4 for 10 is not a 400 hitter, and everyone knows it, because ten at-bats prove almost nothing. Give the same player 500 at-bats and the average starts to mean something real. Claims behave the same way. A handful of them, good or bad, could be a fluke. The carrier waits for volume before it treats your average as the truth about your group.

The rule behind the number: 1,082 claims

Where does Z come from? The oldest answer in the actuarial toolkit is limited-fluctuation credibility, also called classical credibility. It sets a threshold for how much data you need before your own experience can stand entirely on its own. That threshold is 1,082 claims, the count that corresponds to a 90% confidence level and a tolerance of plus or minus 5%.

Hit 1,082 claims in a year and you get full credibility. Your experience carries 100% of the weight. Fall short, and the carrier assigns partial credibility using the square-root rule: Z equals the square root of n divided by 1,082, where n is your claim count. The table below shows how fast that dial climbs, and how slowly.

Read that middle column slowly. To get even half your experience counted, you need roughly 250 claims in a year. A 40-life group generating maybe 100 to 150 claims sits down around 30% to 37%. Two thirds of its result comes from strangers.

One caveat worth stating plainly. The 1,082 figure is the classical textbook standard. Real carriers apply their own credibility tables, often more conservative, and frequently built on covered lives or premium volume rather than raw claim counts, then smoothed over several years so a single year can never swing the price too far. The mechanism is the same. Your year is weighted, not taken at face value.

Why small and mid groups get so little credit

It is tempting to read the table as bureaucratic caution. It is actually math about volatility, and the real driver is dollars, not claim counts.

Health spending is lopsided. A small share of claimants drives most of the cost. One premature infant in the NICU, one transplant, one $2 million cancer course, and a small group's entire year is defined by a single event that had almost nothing to do with how the other 39 families used their coverage. Randomness, not management, wrote that year.

Statistical noise shrinks slowly. It falls with the square root of size, not size itself. Double your group and the swing in your average claims drops by only about 29%, not half. Quadruple it and you cut the swing in half. So a 40-life group's great year could easily be luck, and the carrier has to price as if it might be, because next year the same group could produce the case that blows the doors off.

This is a different mechanism from large-claim pooling, though the two often get tangled together. Pooling caps how much any single catastrophic claim feeds into your rate. Credibility governs how much of your whole experience the carrier believes in the first place. We unpack the pooling side in why healthy groups still get hit with double-digit renewals. Read them together and the renewal stops looking arbitrary.

The community-rating trap: in small group, your experience is invisible by law

Here is the part that surprises most owners. If your company is in the ACA small-group market, your own claims history is not a permitted rating factor at all. Not weighted at 30%. Not weighted at all.

The small-group market generally covers employers with 1 to 50 employees, though four states extend it to 1 to 100 under the PACE Act: California, Colorado, New York, and Vermont. Plans there use modified community rating. Premiums may vary only by age, capped at a 3-to-1 ratio between oldest and youngest, by tobacco use, capped at 1.5 to 1, by geographic area, and by family size. That is the entire menu. You can read the framework straight from the source at the CMS market rating reforms page.

So a healthy 30-life group and a claims-heavy 30-life group of the same ages, in the same county, pay the same community rate. Your great year cannot lower your premium. It just improves the carrier's block. You are subsidizing the pool with no way to bank your own good behavior.

The rule is not there to punish you. Modified community rating exists so a small employer with one sick member cannot be priced out of coverage entirely, which is exactly what happened to many groups before the ACA. The trade-off is symmetric. The same wall that stops a carrier from surcharging your neighbor's cancer year stops you from banking your healthy one. Fairness for the pool, frustration for the employer who ran a clean year.

Cross 51 employees, or 101 in the four expansion states, and everything flips. You land in the large-group market, where plans are experience-rated and credibility finally applies. Now your own history can move your price, for better and for worse. That threshold is a genuine fork in the road, and it interacts with how your workforce ages. We walk through both sides in age-banded vs. community-rated health insurance.

The compounding cost of low credibility

Low credibility would be tolerable if it evened out. It does not, and here is why. The manual rate and medical trend still land on your renewal every single year, while your good years get discounted before they ever reach the price. Bad years feed the manual rate for everyone. Good years get diluted. The ratchet only turns one way, which is part of how family premiums climbed 26% over five years even for employers who managed utilization well.

Walk through an illustrative example. Say a 40-employee group beats its expected claims by $150,000 in a strong year. Real money, earned by a healthy workforce and decent plan design. At roughly 48% credibility, only about $72,000 of that favorable result informs the renewal. The other $78,000 of good news gets blended away into the pool. Then the carrier applies its trend increase on top of the blended base, because trend does not care about your good year. This is illustrative, but the shape is exactly what employers see: a real win, taxed by credibility, then buried under trend.

Notice what that does over a decade. Every rough year across the wider market lifts the manual rate you are partly priced against, whether or not your own group had a rough year. Every good year of your own gets weighted down before it lands. Run that asymmetry across five or ten renewals and it compounds into a base that sits well above where your own claims history alone would put it. The gap is not a billing error. It is the credibility mechanism doing exactly what it was built to do.

Want to pressure-test this against your own numbers rather than a hypothetical? The Premium Renewal Stress Test models how much of a favorable year actually survives into your renewal, and what a bad year would do in reverse.

How to find out how much credibility your renewal actually reflects

You do not have to accept the renewal as a black box. Put three questions to your carrier or broker in writing, and ask for the arithmetic, not a summary.

• What manual rate did you use, and what is our group's own experience? You want both numbers side by side, the pooled baseline and your actual claims.
• What credibility or pooling factor did you apply to our experience this year? This is the Z. If they will not name it, that itself is information.
• Show the blend. Ask exactly how the manual rate and your experience combined into the renewal, so you can see the weighting with your own eyes.

One honest answer you should expect: if you are in the small-group market, the reply to the second question is often "zero, you are community-rated." That is not evasion. It is the law, and knowing it reframes the whole conversation toward the only levers that actually work for you.

What actually earns credit for your own good years

If credibility is the problem, the solutions all point the same direction: get your own experience to count. There are three practical paths.

Grow your covered lives. Credibility rises with the square root of size, so scale genuinely helps, even if slowly. More lives means more claims, a higher Z, and a renewal that leans further on the year you actually had. It is the least controllable lever, but it is real.

Change the funding model so your good years come back to you. This is the direct route, and the market has already moved. In the KFF 2025 survey, 37% of covered workers at firms with 10 to 199 employees are in level-funded plans, and 27% are fully self-funded, against 80% at large firms. Level-funded and self-funded arrangements credit your own experience directly. A quiet claims year can return real money instead of getting averaged into a stranger's bad one. Model the mechanics with the Health Funding Projector, then size the trade-offs and the stop-loss with the Level-Funded Calculator.

Be clear-eyed about the trade. Self-funding hands you the good years and the bad ones, so it belongs with a stop-loss policy to cap catastrophic claims and a cash reserve for the months claims run hot. Level-funding softens that swing with a fixed monthly payment and a potential year-end refund, which is why it has become the on-ramp for groups that want their experience to count without carrying full risk. The right answer depends on your balance sheet's tolerance for a rough month, not just your average year.

Cross into large-group experience rating. At 51-plus employees, or 101-plus in California, Colorado, New York, and Vermont, your history becomes a permitted rating factor again. That cuts both ways, so it rewards a genuinely healthy, well-managed group and punishes a struggling one. Know which you are before you treat the threshold as a finish line.

None of these is a magic switch. The point is simpler than that. Once you understand that credibility is quietly deciding how much of your own story the carrier will hear, you can stop negotiating the trend and start changing the structure that governs the weighting.

Related Reading

• Why Healthy Groups Still Get Hit With Double-Digit Renewals
• When Your Health Plan Has a Good Claims Year: Steps to Use the Leverage
• Age-Banded vs. Community-Rated Health Insurance

Frequently Asked Questions

Why did my health insurance renewal go up after a good claims year?

Because the carrier only partly counts your experience. Your renewal blends your own claims with a pooled manual rate, and a credibility factor decides the mix. For a small or mid-size group that factor is low, so your good year gets diluted while medical trend still applies on top. The result is an increase that looks disconnected from the year you actually had.

What is a credibility factor in health insurance?

It is a weight between 0 and 1 that sets how much of your renewal is based on your own claims versus a pooled baseline built from many groups. A credibility of 1.0 means your experience fully drives the price. A credibility of 0.30 means only 30% of the price reflects your group and 70% reflects the pool. The classical standard reaches full credibility at 1,082 claims in a year.

How many employees do I need before my own claims experience counts?

Two thresholds matter. Statistically, credibility climbs toward full weight as you approach roughly 1,082 claims a year, which usually means several hundred covered lives. Legally, you generally need to exceed 50 employees, or 100 in California, Colorado, New York, and Vermont, to leave the community-rated small-group market and enter experience rating, where your history is allowed to affect the price at all.

Does my company's claims history affect a small-group community-rated plan?

No. In the ACA small-group market, your claims history is not a permitted rating factor. Premiums can vary only by age, tobacco use, geographic area, and family size. A healthy small group and a claims-heavy one of the same ages and location pay the same community rate, so a good year cannot lower your premium.

How can I get credit for a good claims year?

Move to a funding model that returns your experience to you. Level-funded and self-funded plans credit your own claims directly and can refund a favorable year instead of averaging it into a pool. Growing your covered lives raises your credibility over time, and crossing into large-group experience rating lets your history influence the price. Modeling tools can show which path fits your size and risk tolerance.

References

• KFF 2025 Employer Health Benefits Survey
• CMS Market Rating Reforms
• Limited-fluctuation (classical) credibility standard, Casualty Actuarial Society / Society of Actuaries study material.